Pathological Complete Response and Accelerated Drug

Approval in Early Breast Cancer

Tatiana M. Prowell, M.D., and Richard Pazdur, M.D.
N Engl J Med 2012; 366:2438-2441June 28, 2012

New drugs for breast cancer have historically been approved first for patients with metastatic
disease who have few remaining options for systemic treatment. Approval for an adjuvant
indication occurs years later, after large, randomized trials with prolonged follow-up have
been conducted in patients with early-stage disease. Recently, neoadjuvant trials have
introduced new drugs preoperatively in patients with localized breast cancer. Such treatment
aims to render locally advanced cancers operable, facilitate breast-conserving surgery, and
ultimately improve survival. The rate of pathological complete response absence of
residual invasive cancer on pathological evaluation of resected breast specimens and lymph
nodes after preoperative therapy has been used as the primary end point in many
neoadjuvant trials.
Promising investigational drugs should be incorporated into standard treatment for earlystage breast cancer as rapidly as possible to provide the greatest benefit to the most
patients. But this goal must be weighed against the limited safety data available for new
drugs when they are used in patients with curable cancer and uncertainty about whether
improvement in pathological complete response will predict improvements in long-term
disease-free or overall survival.
The uncertainties regarding the risks and benefits of new neoadjuvant drugs may be
managed by enrolling patients who have the greatest risk of recurrence with existing
therapies and are likely to benefit the most. Although modern cytotoxic regimens have
reduced 10-year breast-cancerrelated mortality by approximately one third, certain patients
with early-stage breast cancer, particularly those with high-grade tumors that are negative
for estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor
receptor 2 (HER2) (i.e., triple negative), remain at substantial risk for distant metastatic
disease and death.
Randomized neoadjuvant trials suggest that a pathological complete response may predict
disease-free or overall survival among patients with early-stage breast cancer who are
treated with preoperative systemic therapy. A Cochrane meta-analysis of 5500 patients
enrolled in 14 randomized trials comparing preoperative with postoperative chemotherapy
showed that the risk of death among patients who had a pathological complete response
was about half that of patients with residual tumor at the time of surgery.1
In the United States, regular approval of a new drug requires adequate, well-controlled trials
demonstrating clinical benefit, which is generally defined in early-stage breast cancer as an

improvement in disease-free or overall survival. Alternatively, the Food and Drug

Administration (FDA) may grant accelerated approval on the basis of a surrogate end point
that is reasonably likely to predict clinical benefit. For neoadjuvant breast-cancer treatment,
we propose that the rate of pathological complete response be used as this surrogate. 2 After
accelerated approval, demonstration of an improvement in disease-free or overall survival
would be required; the indication may be withdrawn from product labeling if confirmatory
trials have not shown clinical benefit.
For regulatory purposes, neoadjuvant trials evaluating a new drug with limited safety data
should enroll patients with high-risk features and exclude those with ER- or PR-positive
tumors lacking these characteristics. Patients may be classified as having a high risk of
recurrence on the basis of conventional histologic features or appropriately validated
genomic measures. The highest rates of pathological complete response have generally
been observed among patients with high-grade ER- and PR-negative tumors and those with
HER2-positive tumors.3 Although patients with triple-negative breast cancer have an
increased risk of recurrence, if a pathological complete response is achieved, the likelihood
of survival may be similar to that among patients with more prognostically favorable
subtypes.3,4 Patients with ER- and PR-positive tumors are less likely to have a pathological
complete response to neoadjuvant therapy and more likely to live longer with available
therapy; pathological complete response is thus unlikely to predict clinical benefit in this
subgroup.3 We discourage enrollment of patients with low-grade ER- and PR-positive
tumors into neoadjuvant trials conducted with regulatory intent.
A large, randomized trial of neoadjuvant breast-cancer treatment using an add-on design
studying a standard adjuvant regimen with or without the new drug, all delivered
preoperatively could be used for an accelerated-approval submission. This single
randomized trial, if adequately powered, could both support accelerated approval on the
basis of substantial improvement in the pathological complete response rate and, with
further follow-up, provide data on potential improvements in disease-free and overall survival
to establish clinical benefit. Use of postoperative systemic therapy should be avoided but if
needed (e.g., for completion of a year of adjuvant trastuzumab for HER2-positive breast
cancer) should be consistent in both treatment groups to avoid confounding interpretations
of disease-free and overall survival. Demonstration, with mature data, of a clinically and
statistically significant improvement in disease-free or overall survival would fulfill the
requirements for regular approval and permit continued marketing of the drug for
neoadjuvant use in breast cancer.
At the time of accelerated approval, characterization of long-term toxic effects may be
incomplete, and uncommon adverse events may not be recognized or fully described. A
comprehensive safety assessment is critical in evaluating the benefits of neoadjuvant
therapy for patients with early-stage breast cancer, among whom long-term survival is
common and may result solely from local therapy. There is a risk that a drug approved in this

way could be marketed for a prolonged period, exposing many patients with curable disease
and potentially normal longevity to the risks posed by an ultimately ineffective therapy. To
mitigate this risk, randomized neoadjuvant trials conducted with marketing intent should be
limited to subpopulations at high risk for recurrence despite optimal local and systemic
therapies, and confirmatory trials should be ongoing at the time of accelerated approval.
The trial design described above would isolate the new drug's effect and provide a larger
body of safety data at the time of accelerated approval. Continued follow-up in the same trial
would provide essential information on late or cumulative toxic effects, as well as mature
efficacy outcome data, far more quickly than a subsequent adjuvant trial could do and would
hasten clarification of the relationship between the pathological complete response rate and
survival. For drugs with more extensive prior use in breast cancer or evidence of
unprecedented efficacy, or for drugs being studied in ongoing randomized adjuvant trials,
alternative approaches may be acceptable and should be discussed with the FDA.
The proposed magnitude of the difference between treatment groups in the pathological
complete response rate should be prespecified and have a high likelihood of translating into
a meaningful improvement in disease-free or overall survival; the sample size required to
demonstrate a significant difference in survival may be substantially larger than that needed
to demonstrate a significant difference in pathological complete response rates. Statistical
analyses should use the full intention-to-treat population. Since distant metastatic disease
will develop in some patients with a pathological complete response, small absolute
improvements probably will not have a meaningful effect on long-term clinical benefit;
substantial improvements may be needed to improve disease-free or overall survival. For
example, in a neoadjuvant trial of chemotherapy with or without trastuzumab, the group that
received trastuzumab had a near doubling of the pathological complete response rate (39%
vs. 20%) and a 3-year disease-free survival rate of 71%, as compared with 56% in the other
treatment group.5 Similarly, adjuvant trials of chemotherapy with or without trastuzumab
have demonstrated an approximately 50% relative (12% absolute) improvement in diseasefree survival when trastuzumab was added.6
Despite the promise of pathological complete response as an end point for accelerated
approval, unresolved issues remain, including the definition of such a response that
optimally predicts long-term clinical outcomes, the intrinsic breast-cancer subtypes most
likely to show such a response, and the magnitude of improvement needed to produce
meaningful improvements in disease-free and overall survival. In collaboration with
international investigators, the FDA is conducting a meta-analysis using primary-source data
from more than 12,000 patients enrolled in randomized neoadjuvant trials, aiming to identify
those in whom a pathological complete response is most likely to predict clinical benefit by
correlating this end point with disease-free and overall survival in intrinsic breast-cancer
subtypes.

The FDA has released a draft Guidance to Industry,2 outlining a pathway to accelerated
approval for neoadjuvant breast-cancer therapies and seeking public comments on the use
of pathological complete response as an end point for accelerated approval.

SOURCE INFORMATION
From the Office of Hematology Oncology Products, Food and Drug Administration, Silver
Spring, MD (T.M.P., R.P.); and the Breast Cancer Program, Sidney Kimmel Comprehensive
Cancer Center, Johns Hopkins University, Baltimore (T.M.P.).